Re: [soa-rm-ra] Proposed SCA Event Processing Model

[mpoulin@usa.com]

Well, DDS is a data-centric model while we are working on the service-centric model. 

I would agree with Danny if we clearly and unarguably state in 4.3 that any interaction with the services may happen through its described/announced interfaces only. That is, no object-oriented, event-oriented, data-centric etc. interactions with the service beneath the service interface are allowed to avoid the risk of service coupling in the lower-level service implementation, whatever it may be. 

- Michael

________________________________________


Subject: Re: [soa-rm-ra] Proposed SCA Event Processing Model
	From: Danny Thornton <danny_thornton2@yahoo.com> 
	To: "soa-rm-ra@lists.oasis-open.org" <soa-rm-ra@lists.oasis-open.org>, "Estefan, Jeff A" <jeffrey.a.estefan@jpl.nasa.gov> 
	Date: Wed, 6 May 2009 10:44:00 -0700 (PDT) 
________________________________________

While we are on the pub/sub topic, Data Distribution Service by OMG, http://en.wikipedia.org/wiki/Data_Distribution_Service , is another publish subscribe standard.  DDS has multiple vendors and products and has an efficient multicast capability. DDS and SCA will make for a good contrast.  I do not see that section 4.3, Interacting with Services Model, has any conflicts with DDS or SCA since the RA is an abstraction level above.

Danny